Network Analysis Identifies Mitochondrial Regulation Of

Article

Network Analysis Identiﬁes Mitochondrial Regulation of Epidermal Differentiation by MPZL3 and FDXR

Graphical Abstract Authors Aparna Bhaduri, Alexander Ungewickell, Lisa D. Boxer, Vanessa Lopez-Pajares, Brian J. Zarnegar, Paul A. Khavari

Correspondence [email protected]

In Brief To identify non-transcription factor regulators of epidermal differentiation, Bhaduri et al. devised a network approach called proximity analysis and applied it to identify MPZL3 as a mitochondrial protein required for epidermal differentiation. MPZL3 promotes the enzymatic activity of mitochondrial protein FDXR, and the factors together regulate ROS-mediated epidermal differentiation.

Highlights d Proximity analysis successfully reconstructs expression- based networks in eukaryotes d Proximity analysis identiﬁes MPZL3 as highly connected in epidermal differentiation d MPZL3 is required for differentiation and localizes to the mitochondria d MPZL3 binds FDXR, and together they regulate ROS required for differentiation

Bhaduri et al., 2015, Developmental Cell 35, 444–457 November 23, 2015 ª2015 Elsevier Inc. http://dx.doi.org/10.1016/j.devcel.2015.10.023 Developmental Cell Article

Network Analysis Identiﬁes Mitochondrial Regulation of Epidermal Differentiation by MPZL3 and FDXR

Aparna Bhaduri,1 Alexander Ungewickell,1,2 Lisa D. Boxer,1,3 Vanessa Lopez-Pajares,1 Brian J. Zarnegar,1 and Paul A. Khavari1,4,* 1Program in Epithelial Biology, Stanford University School of Medicine, Stanford, CA 94305, USA 2Division of Hematology, Stanford University, Stanford, CA 94305, USA 3Department of Biology, Stanford University, Stanford, CA 94305, USA 4Veterans Affairs Palo Alto Healthcare System, Palo Alto, CA 94304, USA *Correspondence: [email protected] http://dx.doi.org/10.1016/j.devcel.2015.10.023

SUMMARY differentiation, may be also regulated beyond the nucleus, including in the mitochondria, the lysosome, and the ribosome Current gene expression network approaches (Kasahara and Scorrano, 2014; Settembre et al., 2013; Xu commonly focus on transcription factors (TFs), et al., 2013; Xue and Barna, 2012). Understanding how these biasing network-based discovery efforts away from levels of regulation interact with canonical nuclear gene regula- potentially important non-TF proteins. We developed tion processes can better illuminate how differentiation pro- proximity analysis, a network reconstruction method ceeds through feedback and feedforward loops and may that uses topological constraints of scale-free, provide additional insight into treatments for congenital and somatic diseases of deranged differentiation. small-world biological networks to reconstruct rela- Gene expression network reconstruction is a commonly used tionships in eukaryotic systems, independent of sub- method of describing organism level systems (Herrga˚ rd et al., cellular localization. Proximity analysis identified 2008; Tong et al., 2004), identifying key relationships that can MPZL3 as a highly connected hub that is strongly highlight regulators of a process (Carro et al., 2010), and delin- induced during epidermal differentiation. MPZL3 eating cell types at a transcriptional level (Yang et al., 2013). was essential for normal differentiation, acting down- While the applications of network-based analyses are vast, cur- stream of p63, ZNF750, KLF4, and RCOR1, each of rent network algorithms favor models of less complex organisms which bound near the MPZL3 gene and controlled such as Escherichia coli or networks with a bias toward known its expression. MPZL3 protein localized to mitochon- transcription factors. The recent DREAM5 consortium analysis dria, where it interacted with FDXR, which was of the highest performing, most used network reconstruction itself also found to be essential for differentiation. algorithms highlighted that while combining the outputs from multiple existing network algorithms improved upon the perfor- Together, MPZL3 and FDXR increased reactive mance of a single algorithm alone, the ability to reconstruct oxygen species (ROS) to drive epidermal differentia- known relationships fell significantly from in silico networks to tion. ROS-induced differentiation is dependent upon E. coli networks to eukaryotic Saccharomyces cerevisiae net- promotion of FDXR enzymatic activity by MPZL3. works (Marbach et al., 2012). Additional deconvolution efforts ROS induction by the MPZL3 and FDXR mitochon- aimed to improve these metrics, but were only able to do so in drial proteins is therefore essential for epidermal an incremental manner for eukaryotes (Feizi et al., 2013). There- differentiation. fore, significant challenges persist in using network reconstruction approaches to understand human tissue differentiation, especially when searching beyond transcription factors. INTRODUCTION The epidermis is an excellent model for the application of a network reconstruction approach to discover non-transcription Somatic differentiation requires the concurrent regulation of a factor regulators because it is a relatively well characterized tis- large number of genes that are necessary for the tissue-specific sue with in vitro model systems derived from primary human functions of differentiated cells. Much of this regulation comes cells. The epidermis is comprised of a basal layer of progenitor from transcription factors that regulate gene expression, cells that give rise to the layers of epidermis that exit the cell silencing genes that maintain a proliferative cell-cycle state cycle, enucleate, and provide barrier function through expres- and activating genes that are necessary for the cellular functions sion of tissue-specific differentiation genes. The transcriptional characteristic of the differentiated tissue. However, only a frac- master regulator of the epidermis, p63 (Mills et al., 1999; Truong tion of the genes that are induced during a somatic differentiation et al., 2006; Yang et al., 1999), targets key genes including program are nuclear factors, and recent work suggests that ZNF750 (Boxer et al., 2014; Sen et al., 2012) and MAF/MAFB (Lo- differentiation processes, as well as disease-causing defects in pez-Pajares et al., 2015). Other transcription factors implicated

444 Developmental Cell 35, 444–457, November 23, 2015 ª2015 Elsevier Inc. in the regulation of epidermal differentiation include KLF4 (Patel straints is that by simulating the preferential attachment models et al., 2006; Segre et al., 1999), GRHL3 (Hopkin et al., 2012; Yu that have been described to result in scale-free networks (Bara- et al., 2006), and OVOL1 (Teng et al., 2007). Recent work gener- basi and Albert, 1999), the relationships between genes that are ated kinetic gene expression data in the regeneration of differen- biologically related will be highlighted by higher probabilities tiated epidermal tissue (Lopez-Pajares et al., 2015). The ability to within this network. We first optimized proximity analysis for use this type of kinetic data in a model with well-characterized various input parameters on a series of in silico networks gener- positive controls makes it an ideal system for applying network ated by DreamNetWeaver (Schaffter et al., 2011)(Figure S1C). reconstruction approaches to discover new regulators. We used proximity analysis to analyze the Saccharomyces cer- Here, we develop and apply proximity analysis to network evisiae expression data provided to the DREAM5 network chal- reconstruction during the process of epidermal differentiation. lenge (Table S1) and compared our output to the top-performing Analyzing a time course of gene expression during differentiation algorithms that have generated networks from this data (Feizi generated a network of strongly connected genes, including et al., 2013; Marbach et al., 2012). Because of our interest in those with known roles in differentiation as well as novel candi- identifying non-transcription factor regulators, we used an alter- date regulators. A top hit, MPZL3, is highly induced in the native evaluation metric to the binding site derived metrics used process of epidermal differentiation and downregulated in in the DREAM5 challenge. Using validated gene ontology (GO) cutaneous squamous cell carcinoma. MPZL3 was found to classifications, a frequently used conservative network valida- be essential for epidermal differentiation. Its expression was tion metric (Chagoyen and Pazos, 2010), proximity analysis out- controlled by several known transcriptional regulators of differ- performed the other examined algorithms in both the number entiation, including p63, ZNF750, KLF4, and RCOR1. Live-cell and fraction of edges belonging to the same GO term (Fig- vicinal protein labeling followed by mass spectrometry demon- ure S1D). We additionally noted that while all algorithms exam- strated that MPZL3 primarily interacts with mitochondrial pro- ined had an average local clustering coefficient characteristic teins, with mitochondrial localization confirmed by electron of a small-world network and a power law degree distribution microscopy. Among MPZL3-interacting proteins was FDXR, a characteristic of a scale-free network topology (Figure S1E), mitochondrial enzyme that catalyzes the reduction of ferredoxin. proximity analysis was the most small-world (Figure S1F) with We observed that FDXR is also required for normal epidermal dif- the smallest residual when compared to a perfect power law dis- ferentiation, and its ectopic expression is capable of rescuing the tribution (Figure S1G). These data suggest that in a eukaryotic differentiation defects of MPZL3 depletion. FDXR, which had system, proximity analysis may be well-suited to reconstruct a been previously characterized as necessary for ROS-mediated biologically relevant scale-free, small-world network. apoptosis, was found to control epidermal cell ROS levels in concert with MPZL3, with both proteins mediating ROS-medi- MPZL3 Is a Highly Connected Hub in Epidermis ated epidermal differentiation. MPZL3 and FDXR action in We applied proximity analysis to time course expression data of differentiation is contingent upon FDXR’s enzymatic ability, the process of epidermal differentiation in both confluent differ- demonstrating an intricate role of mitochondrial-based proteins entiation culture conditions as well as an organotypic model of in epidermal differentiation. Taken together, these data develop the skin (Table S2)(Kretz et al., 2013; Kretz et al., 2012; Lopez- a new network construction approach to identify an essential Pajares et al., 2015). To identify genes of interest, we ranked role for MPZL3/FDXR-mediated induction of ROS in epidermal genes in the resulting network by the extent of their small-world differentiation. interaction profile and highlighted genes of potential disease relevance by weighting these hubs by the degree of their change RESULTS in cutaneous squamous cell carcinoma (cSCC) expression datasets (Arron et al., 2011)(Figure 1A), generating a dynamic hub Proximity Analysis score for the most differentially expressed genes (Table S3). To identify regulators of genomic expression in eukaryotes, we We chose this filter for disease relevance because cSCC is a dis- designed proximity analysis, a network-based approach that ease characterized by a loss of differentiation (Kwa et al., 1992; implements topological constraints on a correlation-based Ratushny et al., 2012; Rowe et al., 1992) and candidates high- network. Extensive descriptions of networks in yeast, protein- lighted by this filter were more likely to also be of interest in the protein interaction networks, co-expression networks, and study of cSCC or other differentiation-relevant malignancies in metabolic networks have been characterized by a scale-free, the future. The resulting network of epidermal differentiation small-world network topology (Hughes et al., 2000; Ravasz had the expected scale-free, small-world topology, with many et al., 2002; Tong et al., 2004; van Noort et al., 2004; Yook edges classified in the same GO term (Figure 1B; Figures S2A– et al., 2004). Proximity analysis uses rank-ordered gene expres- S2D). GO term analysis of the top 500 most connected genes sion data followed by the generation of a Pearson correlation from proximity analysis highlighted biological processes related matrix for every pair of expressed genes (Figure S1A). Using to epidermal differentiation (Figure 1C), and nearly all transcrip- this correlation value as a weight, nodes are preferentially con- tion factors known to regulate epidermal differentiation scored nected to a random seed network, with the nodes with more in the top 10% of all scored nodes (Figure S2E), indicating that connections being more likely to gain additional connections proximity analysis successfully highlighted known regulators of (Figure S1B). Each of the edges generated by each simulation interest. is given a probability, and after thousands of simulations, edges The top dynamic hub scoring genes from proximity analysis with the highest probabilities of inclusion are used for down- highlighted known regulators of differentiation, OVOL1 and stream analyses. The hypothesis behind these topological con- RHOV, as well as genes known to be altered in human diseases

Developmental Cell 35, 444–457, November 23, 2015 ª2015 Elsevier Inc. 445 Keratinocyte Differentiation Network A Proximity Analysis Workflow B MPZL3 Hub Identification

Network Generation RNF2

BIVM

SSR1

C6ORF225

LOC440995 LTBP2 EFHA2 WDR45 KIAA1543 LOC285708 SNORD123

RABL5 ADM BAG1 SV2A HSD17B11 C12ORF23 HSD11B2

NEAT1 Other top hits HIST1H2AE EYA2

UGDH

SMARCA1 BAK1P1 ZNF347

L3MBTL SNX7

LOC100144603 SLC46A3 LRPAP1 C12ORF24 SFRP1 GRN UHRF2 SF3B4 IGF2BP1 RCC1 SGCB CYB5D2 CCNI C16ORF7 C20ORF117 CHMP2A LOC284294

TLR5 PEX11B RBMY2FP PTPRU GBA2 KIAA1239 CCPG1

UAP1 Identify connected SLC12A6 LOC153811 Topologically ENDOU SIDT2 GTF2H2B Correlation matrix CTSO LOC646347 KIAA0319L GNL3L LOC147727 NAMPT GLTPD1 LPCAT2 SF3A1 PITPNM3 MBD5 FLJ27352 SLC35B3 PCGF6 KIAA1958 TEAD4 KCNK7 SMPD3 TARDBP SC65 PRDX5 PFDN5 OAS2 [OVOL1, RHOV, RFK TCN2 FLJ20021 ANKRD29 TEF PPP1CC DDX11 AATK SLC44A3 SLC4A3 ZNF17

GPR172A HMGA1

UBA2 GJC1 C20ORF195 FUT2 LRRN4CL MAP2K1 EWSR1 LOC286071 PSMD10 CUL4B SMNDC1 CIDEB IFNAR2 DDX18 CCDC3 RAB7B ERP27 RBM38 ZNF569 RPS23

RBM42 ODZ2 DGKD PDP1

ITM2B LOC100133211 ZNF410 SLC25A38 ZFP106

NADSYN1 TTC27 RCBTB2

NPR2 APOL6 THBD ING4 of keratinocyte constrain by scale- genes by small- TTC3 FAM83C

STT3A

HCLS1

MX1 CARD18,

MEGF6 ZNF195

PSKH1 STX19 C6ORF136 ID2B MANBA TBL1X SEMA5B

SPATA5 C11ORF52 FGF17 TMEM216 SLC16A1

LOC402778 TRIB3 ZNF26

FAM83G TAF1A MMP28 TMPRSS11D

LMCD1 CORO7 LOC100131801 CSNK2A1P RNF40 C11ORF84 UBL5

TPI1P1 WARS EPCAM

IRX3 SYMPK PGAM1 KIAA1715 world interaction ZNF202 free and small-world SERPING1 LONP1 EIF2S2 differentiation IPO7

CIRH1A ZFYVE16 PDE2A MIS12

PGK1

CDK2 ID1

SPTBN5 CRNN, ALOX12B]

NME1 DNMT1

IMP4 RPP30

EFTUD2

PGAP3 NDRG2 DUSP5P ACYP2 PAWR

GJA1 GOLGB1 CSE1L

SPCS1 LMNB2

MYEF2 CAD SNRPF

MAP2 NUDT1

AKT1 PRC1 ALS2CR4 SPC24 E2F1 GABRE FSD1 profile PARP1 parameters SMURF2 timecourse STOML2 BUB1B RFWD3 RNF26 MRPS2 C1ORF51 NUDT11

QTRTD1

CDCA7L DPP9

C3 ASF1B LSM3 WDTC1

CDC25C RAD51AP1 MBOAT7 HSPE1 CNKSR2 C2ORF44

C4ORF46 NEXN

USP13 MAPK3 IKBIP RRP15 HNRNPA3

METTL11A CENPE TRIM11 SH3BGRL3 NUP107 C17ORF59 DCUN1D5 LIX1L MRPL18

TTK FJX1 VSNL1 NUP160 WRAP53 PRPF4 DDX39 RPL39L ABHD10 KLHL13 LTB4R CCDC138

KLHL31 PTHLH CCNA2 KIF20A CCDC99 NOS1 BUD13 MUM1L1 TSPAN33 UPRT NDC80 PARP2 TUSC3

CTPS ASAP3 DKC1 CCNB1 CENPA SHQ1 SCLY ZNF511 NOL10

UBE2C NEIL3

URB2 EIF2AK4 CDK1

DLD BIRC5 MRPL4 CDHR1 GLT8D1 TCHH CYP3A43

RRP9 ETS1 GJD3 LOC100134189

FTSJ1

STK19 WDR18

MZF1 MPV17L2 ADAM8 NOP56 AHCY SFRS7 AQP3 TGFB2 FXYD3

SLC39A7 CDR2 SRM HAUS6 H2AFY

CDKN1A NETO1 TPP2 NRP2 RAC2

NOL9 PLEKHH3 EFNA1

TPRG1L ACVR1B API5

CCNE2 FXYD5 RUVBL1 LYPD5 DHX15 RELL1 PLK4 USP39 SSBP2 FAM45A FAM43A NUDCD1 NCAPG2 LOC647946 ZNF662 C3ORF26 KIFC1 TUBA1B

AKR1C1 RP1

RAGE IFRD2 ZYG11A INF2 ABHD2 LOC100129884 RGL2 CNN3 FAM43B

LIMA1 SKP1

LOC100133632PLA2G4B

MAP7 C4ORF48

PDCD4 ELL3 PARK7 ZNF22

CPNE2 RBM25 AKAP9

AXL PSMA7 NARS2 RAD51L3 LOR

GOLGA7 MGC45800 Disease Relevance GPN3 KRT18P26

CCDC77

CORO1A CDK4 ATRX AADAT LMTK3 MPP7 CENPT PGRMC2 WNT7A CWH43 FOXM1 AURKB FANCG HJURP

FBL PIF1 FAM119A TK1 LAPTM4B MXD1 NLRP1

IPO4 LOC283070 SH3GL1P2 MKI67 GOT2

NUF2

NXPH4 CDCA4

CTNNAL1 KIF23 NIPAL2 ACSL1 DNAJB2

KIFC2 RXRA PLEKHG6

SKP2

ENY2 CRLF1 RNASE4 HS3ST6

CYB5R1

FBXO5 LOC338620

LOC653877 TM7SF2

MAML3 TREX2

PAK1IP1

SFTA1P TMEM14D S100A4 MCRS1 BCCIP DSP NQO1 ABHD11 4−MAR COQ3 CGRRF1 ITSN1

IMPA2

OAT VIPR1 FARSA DAP LOC646050 GPR27 Weight hubs by OSBPL10 CEP170

ESPL1 POLQ TUBB PLA2G4F CDCA8 PHB C1QBP SPHK1 SULT1B1 DUOX2 MGAT2

PTRF MPZL2 DHX33 SCNN1G KIF11 ATXN3

ARHGDIA ANXA5 MAP7D2 DHFR BST1 B3GNT4 PRKCDBP ILK LOC100132308 PPT1 RASSF5 LYPD6B

SLC26A2 UQCR11 HLA−G MTFMT

PLAU GAD1

GNB4 MACROD2 FAM176A

GTPBP6 PERP

PNPT1 Dynamic Hub C1ORF106 MRPL3 CDC25A MTP18 HLA−A

RAB3D

FANK1

CDKN3

TPX2

SSR3 NBEAL2 RTCD1 FLG2 FAM111B C12ORF75

degree of change MICB HINT3 LOC100128526 GM2A CELSR2 ATAD2 GRK6

TRIP13

PLK1 PLEC EMP3 KHNYN LIG1 CREG1 C20ORF46 C11ORF63 PC KIF7 ZBTB7C

FAM162A ZNF610

GSG2 CKAP2L

DHRS11 BNIPL EXOSC6 SLC35B1 NCRNA00185 ARHGAP32

NOG KLK8 PTP4A2 FLVCR1 MRPL1 MXRA5 NDRG4FAM8A1 RAB24 BTG3

GTPBP10 LRRC40 DENND1B ACSL4 PPP2R3B SHC2 VARS2 LOC100133109 RSAD1

RNASEH2A GPRIN2

FASTKD2 KNTC1

RSC1A1 KDM5B

FDX1L PKP1 SCNN1B Score DDAH2 ASAH1 FGD3 LOC285741 ATP5B ODC1 CYP4F3 NDUFA4L2 GPR89A PRMT1 in cancer datasets CHMP1B HIST1H2BG PRIM2 FN1 C7ORF10 FAM96A TRIM73

SHCBP1

CKS1B NCAPD2 CAPN2 RFC3 DUOX1 ZNF467 PTGES GPR146 CYP2U1 MUC1 H2BFS

DSG3 CPE BOC C19ORF10 ALOX15B RAB25 LY6D CMPK2 TPD52L1 KRT13 ANO8 HYAL1 ZNF630

C14ORF145 CRABP2 DEPDC7 RHOV CNGA1 CEP192

CLCA4

EBNA1BP2 DSCR6 SPRR1B

APOO PLS1 LOC144874

USP5

MANSC1 KIF4B C6ORF211 MASTL CLEC2B CDKN2B C11ORF51 ALDH3B2SERPINB13 TINAGL1 C12ORF48 USP28 RHCG HSPBL2 FANCB CHAF1B PIK3C2B OAS3 NOD2SERPINB4 KRT16 ECHDC3 EPHB6 GALC PPL PNRC1 S100A9 GJB6 SERPINB3 TSPAN1 USP18

C15ORF23 ABLIM1 SPRR1A LOC644877 NCF2 AKR1C3 PGLYRP4 BBOX1 SMPDL3A SLC28A3 JHDM1D ZNF750 DEFB1 RAB27B THBS1 CENPH FAM46A USP24 GPR56 ETHE1 CLIC3 ADAMTSL4 HIST1H2ACRAPGEFL1 ABCA12 SPRR2B TMEM79 CASZ1 CSTA MAF DCT ANK3 PVRL4MAFB NLRP10 C10ORF99 CALML5 VEGFC TCN1 TPRG1 STEAP4 RPL32P20 CD274 SULT2B1 SF3A3 ANKRD22 GRHL3 TMPRSS11E KRT77 DHRSX HIST2H2BE CEBPA MCM2 POLR2J3 SPRR3 OVOL1 PPT2 TMEM45B EVPL SLPI RAB9A LYPD3 IL1F9 CAPNS2 HIST1H2BD SP6 KLK11 TMSB15B KIAA0753 ANP32E TKT GPSM1 LAMC2 GRHL1 KLF4

CDC20 TRIM29 GDA DSC1 SPTBN2 MCM7 CRYAB ZNF185 KIAA0513 SPINK5

HMGA2

CYB5B TGM1 C3ORF64 PHF19 C20ORF29 KRT1

MRP63 MMP3 ZDHHC21 LY6G6C NIPAL4 IGFL1 RAET1E MNS1 ATP6AP1 C1ORF54 DUOXA1 NT5E BCO2 LASS3 SPTLC3 C7ORF60 C6ORF132 BTG2 CD55 KRT10 NFIL3 BRCA2 DTYMK ERBB3 POF1B

DLX5 RTN4R SLC9A3

GLRX RASA4 CEP78 C6ORF168 GPR115 TTTY15 COL22A1 ANXA9 EXO1

SLC37A4

SLC39A2 SLAMF7

CEP55

KRT80 RAB34

C20ORF27 ACADVL EPHB3 MX2 KRTDAP CDKN2AIPNL PRRG4 CENPL GLDC C MMACHCGDF11 ATP2A1

THOC4 NCAPG DSG1 ADAP2 PBK BCL2L12 MAP3K8 SPC25 TAF7

TMEM138 SIGIRR ACD GINS2

KLHDC8B EFNA3 KIAA0040

NLRX1 S100P MKNK2 TRMT11 SKA3 HELLS SLC8A1 CRISPLD2 ACOX1 BRIP1 C8ORF73 CEBPB LOC100130776

GSTT1 PRSS2 HK2SMC6

METAP1 KLC1 ARHGAP21 HSPC159 PLEKHA7 RPAP3 SYT14 ZC3HAV1 CERK NPM3 TP53INP2 TEAD1 PKD2 LOC644754 ZBTB5 C9ORF46 NCAPD3 PLCD1 SAAL1

ESPN PAIP2B MBOAT1 RBL1 FAM27B PCDHGA10 Differentiation Network Most Connected GO Terms TMCO4 GALNT14 CSNK1A1 FEN1 ZNF787

LOC145783

PORCN SKA1 ACBD6 RBMX THG1L

VNN1 LEF1 FAHD2A GART

KRTCAP3

VARS GCDH

JUNB NCAPH MSTO2P

LPHN2 DIAPH3 PPME1 TIMM10

BACE2 ZWILCH GIN1 MELK CLDN1

KPNA2 TRAF3IP1

WDHD1 KCTD11 ACER1 PFAS ADSL

TMEM48 NDUFS2 TNNT2 SASH1 LASS4 C10ORF26 VRK1 MAL2 ZDHHC16 CHEK1

SNRNP25 CNTLN PAQR5

POLA1 TPST1 LOC100130547 ITGAE RMI1 WDR76 PROSAPIP1

ZWINT OSBPL2 NDRG1 NFATC3 DEK CXORF57 TLE3 ZNF689

RFC4 ERCC8 DMKN

MSH2

FANCI EGLN3 VASN MED20 HS3ST3A1 MAD2L1 MDC1 LOC254128 EBAG9 RNFT2 GMNN PROCR CDC45 G6PD LOC90110 ORC6L LPPR2 MRPL15

EML1 CEBPD TSLP DSC2 PDLIM1 IFI44L LOC339803FLJ35934

CDCA2 GPN1 HAUS1 C14ORF143

LARS2 FBXL2

MYBL2 C12ORF28

WDYHV1

TBC1D1 LOC645638 SCO1 VLDLR CCND3 C11ORF82 MRPL40

ERCC6L LOC100130663

CREB3L4

LOC100129492 FANCD2 PROS1 PARP11 WDR54 TMEM125

PLEK2 FADD C22ORF23 GGT6 BNIP3L SBSN Epidermis Development PGLYRP3 LDHB PLA2G2F PMFBP1 FAM117A

GNAI2 DLX1

MYO1B

DHX58 PDZK1IP1 TIPIN ITPRIP SPATA20 STIL

ISCA2

BCKDHB MFSD6 C6ORF1

FAM59A GAS7 LAP3 NUDT6

XRCC4

TSPAN4 FAM149B1 NFKBIE EPHX4 FRMD5 SSRP1 ALG8 C7ORF27 ATG4C C1ORF135 MCM10 RBP7 CCNG2 ERAL1

SRI

PLCG2 ANKRD35 CEACAM6 LIPH TSEN15

TEPP HEMK1 HMMR SLCO4C1 TCEAL3 RASD1 CDKL2

USP1 EID2B

POC1A S100A12 NUCKS1 Ectoderm Development C3ORF23 LOC100131138 LGALS7 UCP2

LOC285431 MCM5 CSTB SNW1 LOC727872 ELOVL4 FRMD8

ATIC ZSWIM7 CCDC24 ARRDC4 ZNF300

LAD1 THYN1 RAD51 CLDN20 ANKRD32

C10ORF78 C14ORF80 ST3GAL6 CALCOCO1 FAR2 POP1 MLH1 FNIP1 AK1 PIM3 RUNX1 MICA

ZNF334 ANO1

TYW3

CBS PYCR2 DBF4 CDC6 KDM6B C5ORF46 STIM1

SLC5A1HSPA1B FOXRED1

PCYT1B POLE2 IFIT3 CGN FAM173A TTC9

MCM3 DNAJC10 BIK LRP8 GEMIN6 TMEM45A NDUFAF3 PPARD TPBG GINS1

PRIM1 Epidermal Cell Differentiation C1ORF112 GLTP C16ORF48 C18ORF54 SMARCD1 CEP57 BLM CCDC111

LOC653888 C6ORF142 MXI1 PCSK9

EGLN1

PKMYT1 DFNA5 RAD51C RFC2 KRT222 LOC154761

CDCA5 TATDN1 BANF1

CEP152 IMMP1L TYRO3

COX4NB EXOSC9 DONSON C15ORF59 ACAT1 ITFG2 UTP23 GTF2H4 KRT23 MIA

ADHFE1

ZBTB7B

MRPL41 PSMB8

RBBP8 BTRC RDH12 PAFAH1B3 WFDC12 PCSK6 CLN8 GK SNRPA C19ORF48

TMEM200A DHRS9 C15ORF48 SERPINB9 MND1 HIST1H2BK

KLK7 LOC150759 FRS3 KIF2C Epithelial Cell Differentiation TBC1D3B

CSGALNACT1 PRSS27 YEATS4

BPIL2 KLK12 SOX15 DFFA RMND1 OXNAD1 ZFP57 NEK2 STARD3NL KRT6B S100A7 FUT3 AGRN KIF18A

CENPK C2ORF54 SFRS2 KIAA1468 TPR MLF1IP P704P ORC1L UQCRC1 PITX1

GINS3 CALB2 EXOSC8 RPS6KA1 SLC25A15 CNFN DCTPP1 NOTCH2 IFIT1

WFDC5 USP31 SCRN1 PIM1 ISG15 STX12 ULBP2 ASXL1 HSP90B1 KCTD14 SLC6A14 DIMT1L ABCG4 TOMM40 C5ORF34 LOC100129240

ZNF280B SMC2 ZNF641 THAP10 ELF5 NRARP

SGPP2 SCEL TNFAIP2 HOXA9 NOTCH3 IL1F5 NACAD NCLN FAM64A EPHX3 ID2 COPS6 MUCL1 LRTOMT

HSPA4L SORT1 PRR9 DCLRE1BBARD1 KLK13 MAP1LC3A SLC2A6 APEX1 PPP1R3B SPRR4 SNRNP40 ZDHHC2 CAMK1D A2ML1 BICD2 TOPBP1 HAUS8 PI3 SPRR2G KIAA0831 CEACAM19

HES5 CKAP2 AQP9

LPIN1 GDI2 NET1 TRIM59 IVL TET3 CHP2 UGT8 SLURP1 NUP93 MTFR1 FLG SDR9C7 MYH14

SLC25A40 DNAJB1 DTL LOC283278 IGFL2 TMEM86A Keratinocyte Differentiation LBR CHRNA5 NARF APBB2 POU2F2 NPHP1 CD24HERC5 FABP5L2 PIGN KRT78 GDPD3 CYP4F22 ZBTB8A LOC284952 POU2F3 C4ORF3 TUBB2A IL31RA LLGL2 TRAIP DGAT2

EPS8L1 CXORF38 EPPK1 LOC257396 DZIP1L LOC400043 CASP14 SLC27A4 CHST3 BSPRY TMEM55A WDR67 GPR157 BFAR CLCN3

PAQR8 H2AFV

RP9 PTGER3 PON3 TDRKH CLDN4 LOC652726 LOC643008 CA9 CORT CEP120 PTMAP7 MPZL3 IFI44 MFSD5 IER5 LAMB3 CCDC8

SDF2L1 PAM AGPAT1 RTEL1 ELAC1 AP3B1 EGR2 RBCK1 CXADRP2 EPHA4 ITGA3 N4BP2 ATP6V1A HSPA2

SCAMP2 CRCT1

S100PBP PDIA3 GPX1 CDSN ANKRD28 HSPA5

ZNF643

CEP290

CDK2AP2 TWF2

CCNO CLCN6 PLK2 PSTPIP1

ARHGAP18 Epithelium Development CXADRP1

PDIA4

HYLS1 RARRES1 ANKRD18A IL22RA1 C12ORF56 CA2 CRELD2

KIF14

TFB1M TBC1D15 FAM120C

CCDC41

SOD1 RILPL2

B4GALT6 MMS19 RHEBL1

SDCCAG8 ELF4

MUTYH

TRIM62 FHL1 DGKA

DPP3

WDR26

PRSS22 SPINT1 ZNF382 DMPK

C6ORF15 MAPK9 SH3GL1 CYB5R3 PRMT2 FAM109A FPGS

DNAJC17 PHF6 CD1D

Keratinization DUSP12 RHOQ

RNF169

NT5C3L GABARAPL2 LOC440461SPSB2 GALM

ZC3H12C

TTC31 MUTED

FOXC1 C1ORF198 PSG2 TIMM17B

SCMH1

HOXC4 ISOC2

HYOU1 CCDC18

LOC283481

LOC100128155

SMAD5

COX11

TJAP1 PIGB C9ORF64 Peptide Cross-Linking RHOBTB2

0 5 10 15 -log10(p-value) D E F Proximity Analysis Top Scoring Nodes MPZL3 Expression in Microdissection of Disease Rank Gene Function Differentiation and MPZL3 Expression Human Skin Relevance Cancer Datasets Known regulator Repressor of 1 OVOL1 of epidermal c-myc, deranged 60 1500 Basal Layer differentiation cancer expression Suprabasal Layer Required for 40 2 RHOV GTPase 1000 neural crest family member differentiation 20 Mechanism Unknown 3 MPZL3 500 Unknown 0

Disregulated mRNA Expression 4 CARD18 Unknown expression of MPZL3 Change Fold -20 0 in psoriasis

Undifferentiated Keratincoytes K1 Part of EDC, K10 LOR

Disrupted of mRNA Expression vs. Change Fold MPZL3 5 CRNN marker of in eczema differentiation Mutated in non- Epithelial specific SCC bullous congenital Differentiation 6 ALOX12B with high icthyosiform expression in skin (versus normal skin) erythroderma (versus Subconfluent KC)

G MPZL3 Expression in Human Skin anti-MPZL3 anti-MPZL3 longer exposure control

ColVII DAPI MPZL3

Figure 1. Proximity Analysis Identifies MPZL3 as a Highly Connected Hub in an Epidermal Differentiation Network (A) Proximity analysis workflow consists of a correlation matrix generation that uses topological constraints of scale-free, small-world networks to identify genes that are highly connected, and subsequently cSCC data are used to highlight potential regulators of differentiation. (B) Network depiction of top 10,000 edges from the network. (C) Gene ontology terms for the 500 most connected genes in the differentiation network. (D) Top six candidates that emerge from proximity analysis and their characteristics in terms of dynamic hub score, known function, and disease relevance. (E) Expression of MPZL3 in differentiation datasets (n = 14) when compared to subconfluent keratinocytes and in cSCC (n = 21) when compared to matched normal skin. Mean ± SD is shown. (F) mRNA levels by qPCR from laser capture microscopy separation of basal and suprabasal layers of the skin. (G) Immunofluorescent staining of MPZL3 (red) and collagen VII (green) in normal human skin (left) as well as a non-MPZL3 antibody control stained with collagen VII (right). Scale bars, 15 mM. See also Figures S1 and S2 and Tables S1, S2, and S3.

446 Developmental Cell 35, 444–457, November 23, 2015 ª2015 Elsevier Inc. of abnormal epidermal differentiation, such as CARD18 (psoria- differentiation marker genes in MPZL3 knockdown cells. In sis), CRNN (atopic dermatitis), and ALOX12B (ichthyosis) (Fig- contrast, MPZL3 R99Q and truncation mutants lacking the TM- ure 1D). Of interest in this list of top hits, MPZL3 has no known LC domain did not rescue differentiation. Interestingly, the TM- mechanism, nor has it been characterized in human disease. LC region rescued the differentiation defects of MPZL3 depletion MPZL3 was originally identified via its R99Q mutation in a rough (Figure 2C), suggesting it may mediate key functions of the coat (rc) mouse, characterized by a rough coat, hair loss, ulcer- MPZL3 protein in differentiation. MPZL3’s carboxy-terminal LC ations in the skin, and enlarged sebaceous glands (Cao et al., region is therefore required for MPZL3 action in epidermal 2007). MPZL3 knockout mice recapitulated this phenotype and differentiation. additionally exhibited parakeratosis and increased dermal thick- To characterize the spectrum of genes whose expression is ness (Leiva et al., 2014) as well as decreased body weight and influenced by MPZL3, we performed RNA sequencing on differ- lower levels of blood glucose compared to wild-type or heterozy- entiated keratinocytes that were depleted of MPZL3. MPZL3 gous counterparts (Czyzyk et al., 2013). These phenotypes loss altered the expression of 520 genes compared to control suggest an important role for MPZL3 in the skin as well as sys- (Figure 2D; Table S4), with downregulated genes strongly en- temically. MPZL3 is highly induced over the course of differenti- riched for GO terms related to epidermal differentiation (Fig- ation and downregulated in cSCC (Figure 1E; Figures S2F and ure 2E). The genes regulated by MPZL3 had significantly more S2G). Laser capture microscopy of human skin followed by small-world connections than an equally sized random set of qPCR showed that MPZL3 mRNA is expressed in the suprabasal genes (Figure S2J). Comparison of this downregulated gene layers of the skin at a similar scale to classical markers of the signature to known regulators of differentiation indicated signif- spinous and granular layers (Figure 1F), suggesting its transcript icant overlap with four transcription factors—KLF4, p63, is detectable in a pattern similar to these other epidermal differ- ZNF750, and RCOR1—but not with a negative control ANCR, a entiation markers. To localize expression of MPZL3 protein in tis- long non-coding RNA that is required for maintenance of the pro- sue, we generated MPZL3-specific rabbit polyclonal antibodies. genitor state (Boxer et al., 2014; Kretz et al., 2012; Truong et al., Indirect immunofluorescence of endogenous MPZL3 protein 2006)(Figure 2F). Analysis of the profiles of KLF4, p63, ZNF750, with these antibodies confirmed MPZL3 protein distribution in and RCOR1 interestingly indicated that these genes regulate differentiating layers of intact normal human skin (Figure 1G). the transcript levels of MPZL3, in contrast to 54 other known dif- MPZL3 is thus a highly connected hub in the proximity-anal- ferentiation regulators who displayed less impact on MPZL3 ysis-generated epidermal differentiation network that is strongly expression, including MAF/MAFB, STAU1, GRHL3, TINCR, expressed in differentiating epidermal cells. and ANCR (Figure 2G). Chromatin immunoprecipitation (ChIP) qPCR analysis of a putative regulatory region 12 kb upstream MPZL3 Is Required for Normal Epidermal Differentiation of the MPZL3 transcription start site, identified from analysis of On the basis of its high dynamic hub score in our proximity anal- prior ChIP sequencing (ChIP-seq) studies of these factors (Boxer ysis screen and its expression increase over the course of differ- et al., 2014; Sen et al., 2012), indicated that these four transcrip- entiation, we tested if MPZL3 is required for normal epidermal tion factors have enriched binding at this locus compared to an differentiation by reducing its expression with two distinct small IgG control and a gene desert region (Figure 2H). These data interfering RNAs (siRNAs) in human organotypic epidermal tis- suggest that MPZL3 is a functionally relevant downstream target sue. This setting recapitulates the three-dimensional tissue of these transcription factors in epidermal differentiation. We architecture, gene expression, and protein localization charac- tested this by determining if enforced MPZL3 expression teristic of normal epidermis (Lopez-Pajares et al., 2015). rescued the loss of differentiation gene induction caused by MPZL3-depleted epidermal tissue failed to normally express depletion of p63, ZNF750, KLF4, and RCOR1. In each case, early (K1 and K10) and late (LOR, TGM1) differentiation proteins MPZL3 partially rescued loss (Figure 2I; Figures S2K and S2L), in suprabasal epidermal layers (Figure 2A; Figure S2H), and this confirming it is a functionally relevant downstream target of lack of differentiation gene induction was further confirmed at the these known regulators of differentiation. transcript level for these and additional markers of differentiation (Figure 2B). To test if MPZL3 could itself drive epidermal differen- Mitochondrial Localization of MPZL3 tiation, we expressed it in undifferentiated keratinocyte progen- To explore the mechanism of MPZL3 action, we generated a itor populations. In this context, MPZL3 was capable of inducing fusion of MPZL3 to the mutated promiscuous biotin ligase both early and late differentiation genes (Figure S2I). MPZL3 is BirA*, which has previously been used to perform an unbiased therefore necessary and sufficient for induction of key epidermal interactome analysis (Roux et al., 2012). In this approach, pro- differentiation genes. teins physically proximal to the protein of interest are biotinylated MPZL3 contains a predicted signal sequence at the amino-ter- by BirA* in living cells, purified using streptavidin, then identified minal region of the protein followed by an IGV-like domain that by mass spectrometry (Figure S3A). Using this approach, we contains the ortholog of the R99Q mutation previously identified identified 52 proteins proximal to MPZL3 that were not present in the rc mouse (Cao et al., 2007). MPZL3 also contains a trans- in other BirA* fusion protein datasets (Mellacheruvu et al., membrane (TM) domain and a carboxy-terminal low complexity 2013) nor in the BirA* control experiment. Unexpectedly, 40 (LC) region (Racz et al., 2009). To identify the domains of MPZL3 (77%) of the MPZL3 interactome involved proteins localized to that are functional in epidermal differentiation, we performed mitochondria (Figure 3A). Several of these proteins are known rescue of MPZL3 knockdown with full-length wild-type MPZL3 either to interact physically with each other or to regulate the as well as the MPZL3 R99Q mutant and several MPZL3 trunca- same mitochondrial pathways, suggesting that MPZL3 differen- tion mutants. Full-length MPZL3 largely restored expression of tiation effects may occur in concert with other mitochondrial

Developmental Cell 35, 444–457, November 23, 2015 ª2015 Elsevier Inc. 447 A Loss of Tissue Differentiation with MPZL3 Depletion B C MPZL3 Rescue with Mutants MPZL3 Knock Down in Organotypic Tissue N IGV-Like TM C Signal Low siCtrl Sequence R99Q Complexity KRT1 KRT10 LOR TGM1 siCtrl siMPZL3Rescue construct ColVII ColVII ColVII ColVII 1.5 siMPZL3 1 + EV DAPI DAPI DAPI DAPI siMPZL3 2 MPZL3 Rescue with Mutants 6 +MPZL3 WT 5 4 +MPZL3 R99Q 1.0 3 siMPZL3 1 2.5 +IGV-Like-TM +TM-LC KRT1 KRT10 LOR TGM1 2.0 ColVII ColVII ColVII ColVII DAPI DAPI DAPI DAPI 0.5 1.5 1.0 siCtrl level Relative mRNA Expression mRNA Relative

Relative mRNA Expression 0.5 0.0 0.0 siMPZL3 2 KRT1 KRT10 LOR TGM1 1 0 R G 1 M 0 ColVII ColVII ColVII ColVII LO FL K1 PZL3 KRT TG K1 LOR FLG M KRT1 SPRR3 PZL3 TGM1 DAPI DAPI DAPI DAPI M SPRR3

D E F Down Regulated Gene Signature Overlap RNA-Seq of MPZL3 KD MPZL3 Down Regulated Genes GO Terms in Differentiation (328 genes) MPZL3 Fold Change Keratinization 175*** KLF4 Targets

siCtrl Epidermis Development siMPZL3 siMPZL31 2 147*** ZNF750 Targets 0.00 Keratinocyte Differentiation

Ectoderm Development 119*** TP63 Targets -5.00 # of genes log (fold change) Epidermal Cell Differentiation 2 85*** RCOR1 Targets Epithelium Development 328 down 3 ns ANCR Targets 0 0 0 10 2 3 -log (p-value)

G H I MPZL3 Rescue of Epidermal Regulators MPZL3 Expression with ChIP qPCR KD of Epidermal Regulators 0.15 IgG KRT1 520 genes TP63 ZNF750 KRT10 5.00 0.10 KLF4 LOR RCOR1 0.00 STAU1i ZNF750i MAF/MAFBi KLF4i TP63i RCOR1i GRHL3i TINCRi ANCRi % of input FLG MPZL3 0.05 -5.00 SPRR3 log (fold change) 192 up 2 5.00 0.00 0.00 TGM1 0.00 -2.00 log (fold change) p63i 2 KLF4i -5.00 gene desert RCORi MPZL3 -12 kb ZNF750i log2(fold change) p63i + MPZL3 KLF4i+ MPZL3 NF750i+ MPZL3 Z RCOR1i+ MPZL3

Figure 2. MPZL3 Is Required for Normal Epidermal Differentiation (A) Knockdown of MPZL3 with two distinct siRNAs in an organotypic model of differentiation indicates that markers of differentiation (K1, K10, LOR, TGM1) are not expressed as highly in MPZL3i samples compared to a control siRNA. Scale bars, 20 mM. (B) Quantification of transcript levels of a panel of epidermal differentiation markers with either control siRNA or MPZL3 siRNA mediated knockdown (KD). n = 3 biological replicates; mean ± SEM is shown. (C) Rescue experiments of the MPZL3 KD phenotype was evaluated in the context of forced expression of wild-type MPZL3, MPZL3 R99Q, or two truncated versions of MPZL3 with siRNAs to KD MPZL3. Values shown are the mean of three biological replications, averaged between two siRNAs. (D) RNA sequencing was performed on keratinocytes that were subjected to differentiation conditions in culture. The heatmap shows genes that are changed at least 4-fold in either of the MPZL3 KD conditions compared to control siRNA. (E) Gene ontology analysis of the genes that are downregulated in the RNA sequencing dataset with MPZL3 KD. (F) Overlap of the genes that are downregulated with MPZL3 KD with the gene signatures of known regulators of epidermal differentiation. The left column indicates the number of genes involved in the overlap and *** indicates a p value < 0.001 (Fisher’s exact test); ns indicates a non-significant p value. (G) MPZL3 expression level in datasets with perturbation of known regulators of epidermal differentiation (Boxer et al., 2014; Hopkin et al., 2012; Kretz et al., 2013; Lopez-Pajares et al., 2015; Truong et al., 2006). (H) ChIP qPCR analysis of genomic regions surrounding MPZL3 genomic locus indicates known regulators of MPZL3 transcript expression are enriched for binding in a proximal genomic region compared to an IgG control. n = 2 biological replicates; mean ± SD is shown. (I) Quantification of transcript levels in heatmap format of a panel of epidermal differentiation markers with the KD of known regulators of epidermal differentiation, with or without forced expression of MPZL3. See also Figure S2 and Table S4. proteins, though none of these have yet been described top 10 interactions were all more than 95% likely to be real inter- as important in epidermal differentiation. To further validate actions, and interestingly these proteins were enriched for mito- the localization of MPZL3, we performed cell fractionation chondrial matrix localization (Figure 3D). (Figure S3B) and electron microscopy in keratinocytes that To better understand the role of MPZL3 in mitochondrial confirmed the majority of MPZL3 localizes to the mitochondria biology, we assayed the mitochondrial morphology of cells (Figures 3B and 3C). SAINT score analysis, a metric for the likeli- depleted of MPZL3 by siRNA compared to control siRNA hood of a true interaction (Choi et al., 2011), indicated that the treated cells. Surprisingly, cells with knockdown of MPZL3 had

448 Developmental Cell 35, 444–457, November 23, 2015 ª2015 Elsevier Inc. A B C Electron Microscopy of Quantitation of MPZL3 MPZL3 Proximal Proteins MPZL3 Mitochondrial Localization Localization in EM Analysis

SARS2 DARS2 LARS2 Mitochondrial ACSF3 12.8% Nuclear Cytoplasmic YARS2 AARS2 RARS2 21.1% Unknown ALDH6A1 54.7% ALDH2 METTL15 KIAA0564 GADD45 GIP1 11.4% ALDH1B1 NDUSFS1 FDXR FLJ95242 EFG ALDH9A1 PFKFBK2 ACADS NFS1

AK4 MRPS22 ACAD9 IDH3G SDHA D E RG9MTD1 TRIM29 Top Mass Spectroscopy Hits Proximity Ligation Analysis of MPZL3 and FDXR MPZL3 Gene Spectral SAINT Sub- siCtrl siMPZL3 siFDXR PNPT1 Counts Score Localization NLRX1 SUCLG2 NDUFV1 GFM2 TREX2 FDXR22 1 Matrix KIAA056420 1 N/A PPOX NDUFA10 MPST MIPEP PMPCB ALDH4A1 GUK1 OASL ACAD917 1 Complex I ALDH4A113 1 Matrix

CPT2 TST TBRG4 GLS NDUFS112 1 Inner Mem DAPI PLA NUBPL CECR5 CLPX Mitochondrial Nuclear TST7 1 Matrix F Cytoplasmic YARS2 12 0.99 Matrix Co-Immunoprecipitation HK2 ACSF2 PMPCA PTCD3 Unknown Vesicular SARS214 0.98 Matrix Input IgG IP SUCLG214 0.96 Matrix IP: MPZL3 anti-FDXR Known Metabolic Interaction Known Physical Interaction Both DARS2 14 0.96 Matrix IP: FDXR anti-MPZL3

GHProximity Ligation Analysis of FDXR with Mutant and WT MPZL3 PLA Quantitation of WT and Mutant MPZL3 with FDXR 5 MPZL3 MPZL3 R99Q IGV-Like-TM TM-LC

1 Number of Interactions/nuclei 0

TM-LC MPZL3 W MPZL3 IGV-Like-TMR99Q

Figure 3. MPZL3 Localizes in Mitochondria and Interacts with FDXR (A) Depiction of MPZL3 proximal proteins identified by mass spectrometry of biotinylated proteins isolated from differentiated cells expressing an MPZL3-BirA* fusion construct. The 53 proteins shown in this network were detected exclusively in the MPZL3-BirA* expressing cells as opposed to BirA* only expressing cells and were filtered for proteins that are commonly found in BirA* experiments. (B) Electron microscopy in keratinocytes expressing an MPZL3 overexpression construct. Antibodies to HA were used to identify MPZL3, and 5-nm gold conjugate was used for detection. A representative image of a mitochondria dotted with gold particles is shown. (C) Quantification of observed localization of gold conjugates tagging the expressed MPZL3-FHH protein. (D) Table representing the top proximal hits from the MPZL3-BirA* tagging experiment indicating number of spectral counts for each hit as well as the SAINT score. (E) Proximity ligation analysis with endogenous MPZL3 and FDXR. The signal representing the interaction is shown in red, and nuclei are marked by the blue DAPI signal. siRNAs targeting either MPZL3 or FDXR ablate the observed signal. Representative images are shown. Scale bar, 7 mM. (F) Co-immunoprecipitation between MPZL3 and FDXR; 1% of input is shown. (G) Proximity ligation analysis on HA-tagged constructs of wild-type MPZL3, MPZL3 R99Q, and two truncations of MPZL3 with endogenous FDXR antibody. The signal representing the interaction is shown in red, and nuclei are marked by the blue DAPI signal. Representative images are shown. Scale bar, 7 mM. (H) Quantification of the proximity ligation analysis signal shown in (G). n = 3 biological replicates with at least 10 images analyzed per replicate; mean ± SEM is shown. See also Figures S3 and S4 and Table S5. significantly smaller mitochondria, though the mitochondrial in these cells. Few of these compounds were changed signifi- membrane potential, as measured by JC-1 aggregation, was cantly in even one of the siRNA experiments, but those that not abolished (Figures S3C–S3F). We additionally stained for were changed were related to pathways including lipid, lipoate, mitochondrial protein, mitofilin, in normal human skin to validate and fatty acid metabolism (Figures S3H and S3I), which is that mitochondria are abundantly present in layers of the skin consistent with the previously identified role of MPZL3 in organ- that overlap with MPZL3 expression (Figure S3G). We also per- ismal level metabolism changes (Czyzyk et al., 2013). formed metabolomics analysis of differentiated keratinocytes The most proximal protein to MPZL3 by both spectral counts with 5 types of mass spectrometry identifying 621 compounds and SAINT score was ferredoxin reductase (FDXR), an enzyme

Developmental Cell 35, 444–457, November 23, 2015 ª2015 Elsevier Inc. 449 that catalyzes the reduction of ferredoxin with an electron from FDXR is highly expressed. In contrast, enforced MPZL3 expres- NADPH and has been studied for its role as a p53 target gene sion could not rescue the differentiation gene loss seen with required for reactive oxygen species (ROS)-mediated apoptosis FDXR depletion (Figure S4G), indicating that MPZL3 pro-differ- via its interaction with Fhit (Hwang et al., 2001; Liu and Chen, entiation impacts depend upon the presence of FDXR. FDXR is 2002; Pichiorri et al., 2009). siRNA mediated depletion of the therefore necessary and sufficient to induce differentiation next five top mitochondrial hits by SAINT score did not pheno- genes in a manner that appears functionally downstream of copy the MPZL3 differentiation defect (Figure S4A), so we MPZL3. focused upon FDXR as a functional candidate to explain MPZL3’s role in epidermal differentiation. FDXR, whose expres- MPZL3 and FDXR Regulate ROS sion was observed to be generally constant during the course of We next explored how MPZL3 and FDXR enable epidermal dif- differentiation (Figures S4B–S4D), therefore represents a candi- ferentiation. To do this, we focused on their roles in generating date partner for MPZL3 function during this process. ROS. A recent study implicated a potential role for mitochondria To characterize the MPZL3-FDXR interaction, we first used in epidermal differentiation via a requirement for intactness of the proximity ligation analysis (PLA) (So¨ derberg et al., 2008) to vali- mitochondrial electron transport chain and ROS production (Ha- date the proximal co-localization of FDXR and MPZL3 in manaka et al., 2013). Additionally, other prior work had observed keratinocytes (Figure 3E; Figure S4E). We next validated the that FDXR influenced ROS-mediated apoptosis in epithelial can- interaction between FDXR and MPZL3 using co-immunoprecip- cer cell lines (Liu and Chen, 2002). We thus tested if the effects of itations, with each protein pulling down the other and further MPZL3 and FDXR in epidermal differentiation might involve im- confirming their interaction (Figure 3F). We additionally per- pacts on levels of ROS in epidermal cells. We first validated formed PLA between endogenous FDXR and an HA tagged that there is indeed an upregulation of ROS in our model of ker- version of wild-type MPZL3, MPZL3 R99Q, the IGV-like TM trun- atinocyte differentiation (Figure 5A). We next noted that deple- cation, and a TM-LC truncation of MPZL3. Interestingly, the tion of MPZL3 and FDXR, alone or together, decreased the functionally active wild-type MPZL3 and TM-LC truncation detectable ROS in these cells (Figure 5B). Moreover, enforced yielded detectable interaction signal with FDXR. In contrast, expression of MPZL3 and FDXR, alone or together, increased the non-functional R99Q MPZL3 and IGV-like TM MPZL3 trunca- the detectable levels of intracellular ROS (Figure 5C), suggesting tion mutants did not (Figures 3G and 3H; Figures S4H and S4I), that the observed differentiation phenotypes with depletion or indicating that FDXR binding is associated with functional forced expression of MPZL3/FDXR may indeed be a result of MPZL3. These findings indicate that MPZL3 protein resides in changes in ROS. mitochondria and raise the possibility of a potentially functional To test this hypothesis, we differentiated keratinocytes in the role for MPZL3 binding to the FDXR mitochondrial protein. context of MPZL3 and FDXR knockdown and treated these cells with either a DMSO vehicle control or with galactose oxidase FDXR in Epidermal Differentiation (GAO), an enzyme that exogenously generates ROS within the If FDXR is a functionally important MPZL3-interacting protein cell. Consistent with the hypothesis that MPZL3/FDXR regulate in the latter’s pro-differentiation impacts, then its loss would differentiation through ROS regulation, GAO treatment rescued also be predicted to impair differentiation. To test this, we the transcript level differentiation defects of MPZL3/FDXR depleted FDXR from organotypic human epidermal tissue and depletion (Figure 5D; Figure S5A). To further confirm this finding, examined the expression of differentiation genes. FDXR deple- undifferentiated cells ectopically expressing MPZL3 and FDXR tion decreased differentiation gene expression, as confirmed at were treated with EUK134, a chemical sponge for ROS, to see the protein and mRNA levels (Figures 4A and 4B; Figure S4F). if blocking MPZL3/FDXR-induced increase in ROS would block These results recapitulated MPZL3 loss. To further understand their pro-differentiation effects. EUK134 reversed the increased FDXR’s role in differentiation, we performed RNA sequencing differentiation marker expression driven by enforced expression in the context of FDXR knockdown in keratinocytes undergoing of MPZL3 and FDXR (Figures 5E and 5F; Figure S5B). Thus, the differentiation induction in vitro. 774 genes were changed, and pro-differentiation effects of MPZL3 and FDXR depend on ROS. overlapping the genes changed with FDXR depletion and those Previous work suggested that ROS in the context of epidermal altered by MPZL3 depletion yielded a significant overlap of 113 differentiation may function via modulation of NOTCH signaling, genes (p value < 0.05) (Figures 4C and 4D). Analysis of these and based upon the regulation of ROS by MPZL3 and FDXR, we genes indicated that they are enriched for GO terms relating to performed gene set enrichment analysis (GSEA) (Mootha et al., keratinocyte differentiation (Figure 4E), suggesting that FDXR 2003) on the RNA sequencing profiles of MPZL3 and FDXR (Ha- may regulate epidermal differentiation in a manner similar to manaka et al., 2013). Indeed, these datasets are significantly en- that of MPZL3. Indeed, similar to MPZL3, ectopic expression riched for genes that are transcriptionally regulated by NOTCH of FDXR in undifferentiated keratinocytes increased the expres- (Figure S5C), and qPCR of selected NOTCH target genes verified sion of markers of epidermal differentiation (Figure 4F), indicating that double knockdown of MPZL3 and FDXR results in the deple- that FDXR is also necessary and sufficient for this process. We tion of several of these targets (Figure S5D). further examined the relationship between MPZL3 and FDXR We furthermore tested the relationship of ROS and epidermal by performing a functional rescue experiment with forced differentiation by overexpressing SOD1 and SOD2 superoxide FDXR expression in the context of MPZL3 depletion. FDXR dismutase genes in the context of epidermal differentiation. expression was sufficient to rescue differentiation defects of Overexpression of these genes reduced ROS levels and the MPZL3 loss (Figure 4G), indicating that, in addition to binding levels of differentiation marker gene expression (Figures S5E MPZL3, FDXR can compensate for MPZL3 function when and S5F). We additionally titrated the levels of EUK134 in

450 Developmental Cell 35, 444–457, November 23, 2015 ª2015 Elsevier Inc. A Loss of Tissue Differentiation with FDXR Depletion B

siCtrl FDXR Knock Down in Organotypic Tissue K1 K10 LOR TGM1 ColVII ColVII ColVII ColVII 2.5 DAPI DAPI DAPI DAPI siCtrl 2.0 siFDXR 1 siFDXR 2 1.5 siFDXR 1 TGM1 K1 K10 LOR 1.0 ColVII ColVII ColVII ColVII DAPI DAPI DAPI DAPI 0.5

Relative mRNA Expression mRNA Relative 0.0 3 0 1 K1 siFDXR 2 K1 LOR FLG FDXR TGM K1 K10 LOR TGM1 MPZL SPRR3 ColVII ColVII ColVII ColVII DAPI DAPI DAPI DAPI

C D E RNA-Seq of FDXR KD Overlap of MPZL3 and MPZL3 and FDXR Geneset in Differentiation FDXR Regulated Genes Overlap GO Terms MPZL3 FDXR

siCtrl siFDXRsiFDXR 1 2 Ectoderm Development Epidermis Development Keratinocyte Differentiation

407 genes Epidermal Cell Differentiation 113* 661 genes 466 down Epithelial Cell Differentiation Keratinization

0 1 2 3 4 5 -log (p-value) p-value < 0.05 10 FG FDXR Overexpression in FDXR rescue of MPZL3 774 genes 150 Differentiation Defects Undifferentiated Keratinocytes 100 300 50 siCtrl + EV EV siMPZL3 + EV 200 4 FDXR OE siMPZL3 + FDXR 100 3 siCtrl + FDXR 308 up 10 8 2 5.00 6 4 1 0.00 2 Relative mRNA Expression mRNA Relative 0 Expression mRNA Relative - 5.00 R 0 log (fold change) LOR FLG 2 FDX KRT1 TGM1 0 MPZL3 KRT10 SPRR3 3 R R G 3 DX LO FL F KRT1 KRT1 TGM1 MPZL SPRR

Figure 4. FDXR Phenocopies and Rescues MPZL3 Differentiation Defects (A) Knockdown of FDXR with two distinct siRNAs in human organotypic epidermis indicates that markers of differentiation (K1, K10, LOR, TGM1) are inhibited in FDXR KD samples compared to control siRNA. Scale bars, 20 mM. (B) Quantification of transcript levels of a panel of epidermal differentiation markers with either control or FDXR siRNA mediated KD. n = 3 biological replicates; mean ± SEM is shown. (C) RNA sequencing was performed on keratinocytes subjected to differentiation conditions in culture. The heatmap shows genes that are changed at least 4-fold in either of the FDXR KD conditions compared to control siRNA. (D) Overlap between genes changed with MPZL3 KD or FDXR KD. The overlap of 113 genes is significant with a p value < 0.05, Fisher’s exact test. (E) Gene ontology analysis of the set of overlapped genes. (F) Quantitation of transcript levels of a panel of differentiation markers in subconfluent cells with enforced expression of FDXR. (G) Rescue experiments of the MPZL3 KD phenotype in the context of forced expression of FDXR with siRNAs to KD MPZL3. Quantitation of transcript levels of a panel of epidermal differentiation markers is shown. n = 3 biological replicates; mean ± SEM is shown. See also Table S6. differentiated keratinocytes and saw a proportional decrease GAO and titrating the levels of galactose resulted in an increase in epidermal differentiation gene marker expression with the of both epidermal differentiation gene expression and ROS (Fig- decrease of ROS (Figures S5G and S5H). Additionally, the corol- ures S5I and S5J). These data agree with previous observations lary experiment of treating undifferentiated keratinocytes with of ROS function and suggest that MPZL3 and FDXR regulate

Developmental Cell 35, 444–457, November 23, 2015 ª2015 Elsevier Inc. 451 AB C

(legend on next page) 452 Developmental Cell 35, 444–457, November 23, 2015 ª2015 Elsevier Inc. differentiation via modulation of ROS, a key signaling molecule where known dominant regulators of differentiation, including required for epidermal differentiation. p63, ZNF750, KLF4, and RCOR1, are necessary to induce We next examined the role of FDXR catalytic activity in this MPZL3 expression, following which MPZL3 binds to stably ex- process. FDXR’s rescue of impaired differentiation due to pressed FDXR, promoting FDXR’s enzymatic activity to generate MPZL3 depletion, along with overexpressed FDXR’s capacity ROS-driven epidermal differentiation (Figure 6E). to induce differentiation in undifferentiated epidermal cells that lack significant MPZL3, suggested that FDXR-driven differentia- DISCUSSION tion may require its enzymatic activity and that MPZL3 may help enhance that activity. To test this possibility, we measured Here, we present a new network approach—proximity anal- NADPH, a substrate of FDXR (Figure S6A), from whole-cell ysis—that is unbiased toward protein subcellular location and lysates of keratinocytes in differentiation conditions. Active that discovered new regulatory mechanisms for mitochondrial FDXR catalytic activity should decrease NADPH levels. As ex- proteins in epidermal differentiation. Proximity analysis uses to- pected, forced expression of FDXR decreased the amount of pological constraints to identify regulators in eukaryotic systems NADPH in the cell. Consistent with this, in differentiation condi- by using biologically influenced network simulation to generate a tions where MPZL3 is highly expressed, MPZL3 depletion probabilistic network that reconstructs network topology, high- increased the measurable levels of cellular NADPH back to levels lighting highly connected nodes. Using proximity analysis, we comparable to control as well as FDXR depletion did, consistent identified MPZL3 as a candidate regulator of epidermal differen- with MPZL3’s impact on FDXR catalytic activity (Figure 6A; Fig- tiation. MPZL3 is required for normal epidermal differentiation ures S6B and S6C). Upon depletion of MPZL3 or FDXR in the and is downstream of known transcription factor regulators of context of enforced FDXR expression, ROS levels were also epidermal differentiation. Mass spectrometry identified that the diminished (Figure S6D). Testing the hypothesis that the relation- majority of proteins proximal to MPZL3 are those known to ship between NADPH and ROS is important in the context of reside in the mitochondria, with electron microscopy validating epidermal differentiation, we ectopically expressed NOX5, an MPZL3’s mitochondrial localization. Interaction with the top hit, NADPH oxidase protein. This experiment increased the levels the FDXR mitochondrial protein, was confirmed by PLA and of epidermal differentiation gene expression, lowered the relative co-immunoprecipitation, and FDXR was also discovered to be levels of intracellular NADPH, and increased ROS within these essential for epidermal differentiation. FDXR rescues the differ- cells (Figures S6E–S6H). These data support the model that entiation defects seen with MPZL3 depletion, and MPZL3 and NADPH and ROS are importantly linked in the context of FDXR together induce ROS, which is required for epidermal epidermal differentiation, highlighting the importance of FDXR’s differentiation. catalytic activity during this process. Previous work on network-based approaches to understand- To further interrogate the role of FDXR catalytic activity, we ing somatic tissue differentiation has primarily focused upon generated mutants of FDXR lacking residues needed to bind transcription factors (Suzuki et al., 2009; Yosef et al., 2013). NADPH and FAD, a cofactor of FDXR (Lambeth and Kamin, Understanding transcriptional regulation of somatic tissue differ- 1976; UniProt Consortium, 2015). Each of these mutants dis- entiation is essential; however, the ability to detect non-tran- played lower levels of enzymatic activity compared to wild- scription factor regulators is an exciting challenge as the frontiers type FDXR, in spite of retaining their binding to MPZL3 (Figure 6B; of understanding biological regulation expand to other parts of Figures S6I and S6J). Unlike wild-type FDXR, none of these mu- the cell. The mitochondria has become a fascinating example tants were able to induce differentiation gene expression in un- of how non-transcription factor regulation can happen in a differentiated keratinocytes (Figure 6C), and the NADPH and myriad of ways, regulating key cellular functions such as FAD binding mutants also failed to fully rescue full expression apoptosis and metabolism (Kasahara and Scorrano, 2014). The of selected differentiation markers (Figure 6D; Figures S6K and epidermis is a tractable setting to seek a better understanding S6L). These data indicate that FDXR enzymatic activity is of this regulation as recent work has extensively explored the required for its function in differentiation and that its enzymatic transcriptional modules and signatures that are characteristic activity is dependent upon MPZL3. Taken together, these obser- of the phases of differentiation (Lopez-Pajares et al., 2015), vations are consistent with a model of epidermal differentiation and smaller scale attempts have been made at exploring the

Figure 5. MPZL3 and FDXR Regulate and Are Required for ROS Mediated Differentiation (A) Reactive oxygen species (ROS) levels in differentiating cultured keratinocytes as measured by 20,70-dichlorofluorescin diacetate (DCF-DA) fluorescence. Hydrogen peroxide (H202) was used as a measurement control. * indicates p < 0.05. n = 5 biological replicates; mean ± SEM is shown. (B) ROS levels with MPZL3 and/or FDXR KD as measured by DCF-DA fluorescence. * indicates p < 0.05. n = 5 biological replicates; mean ± SEM is shown. (C) ROS levels with forced MPZL3 and/or FDXR expression as measured by DCF-DA fluorescence. * indicates p < 0.05. n = 5 biological replicates; mean ± SEM is shown. (D) ROS levels when cells are treated with either DMSO or with GAO as measured by DCF-DA fluorescence. (E) Quantitation of transcript levels of a panel of differentiation markers with KD of MPZL3 and FDXR by siRNA with either DMSO treatment or treatment with galactose oxidase (GAO). n = 3 biological replicates; mean ± SD is shown. * indicates p < 0.05, ** indicates p < 0.01, and *** indicates p < 0.001. (F) ROS levels when cells are treated with either DMSO or with EUK134 as measured by DCF-DA fluorescence. (G) Quantitation of transcript levels of a panel of differentiation markers with enforced expression of MPZL3 and FDXR with either DMSO treatment or treatment with EUK134. n = 3 biological replicates; mean ± SD is shown. (H) Summary heatmap of the ROS rescue experiments. See also Figure S5.

Developmental Cell 35, 444–457, November 23, 2015 ª2015 Elsevier Inc. 453 AB C

MPZL3 KD rescue of Mutant FDXR NADPH Usage FDXR Mutants in Undifferentiated Keratinocytes FDXR NADPH Usage 2.0 2.0 300 Full Length 200 100 E240A 1.5 1.5 * G405A 15 228-229 Deletion W398A 1.0 EV level 1.0 EV level 10 * 406-407 Deletion

0.5 ** *** 0.5 5 * Relative mRNA Expression Relative Amount of NADPH Relative Amount of NADPH ** *** **** ** *** **** ** *** 0.0 0.0 0 h K1 E240AG405A W398A K10 LOR FLG TGM1 FDXR OE Full Lengt SPRR3 EV + siFDXR 228-229 Deletion406-407 Deletion FDXR OE + siFDXR FDXR OE + siMPZL3 NADPH FAD Binding Mutants Binding Mutants

DEFDXR Mutants Rescue of MPZL3 Differentiation Defects K1 FDXR

K10 MPZL3

LOR 3.00 TP63 ROS FLG ZNF750 Mitochondria SPRR3 KLF4 -3.00 RCOR1 log (fold change) TGM1 2 MPZL3 Terminal Epidermal Differentiation

E240A G405A W398A

Full Length

siMPZL3 + EV 228-229 Deletion 406-407 Deletion

Figure 6. FDXR Enzymatic Activity Is Required for Its Role in Differentiation (A) Relative amount of NADPH as measured by fluorescence compared to differentiated keratinocytes treated with empty vector (EV). n = 3 biological replicates; mean ± SEM is shown. (B) Relative amount of NADPH as measured by fluorescence compared to differentiated keratinocytes treated with EV. Mutant FDXR constructs marked in red are predicted to perturb NADPH binding by FDXR, and those in blue are predicted to perturb FAD binding by FDXR. n = 3 biological replicates; mean ± SEM is shown. (C) Quantitation of transcript levels of a panel of differentiation markers in subconfluent keratinocytes subjected to forced expression of wild-type or mutant FDXR constructs relative to EV. n = 3 biological replicates; mean ± SEM is shown. (D) Rescue experiments of the MPZL3 KD phenotype were evaluated in the context of forced expression of FDXR or FDXR mutants with siRNAs to KD MPZL3. Quantitation of transcript levels of a panel of epidermal differentiation markers is shown in heatmap format. n = 3 biological replicates; mean ± SD is shown. (E) Model of transcriptional regulation of MPZL3 by known regulators of epidermal differentiation, followed by mitochondrial localization, interaction with FDXR, and joint regulation of ROS required for terminal epidermal differentiation. See also Figure S6. spatiotemporal relationships that are key in regulating the known roles in epidermal disease, this raises the possibility epidermis from a network-based perspective (Grabe et al., that other candidates from this list, including MPZL3 itself, may 2007). Proximity analysis highlighted many known regulators of have a role in the pathogenesis of cSCC or other epithelial epidermal differentiation, as well as a host of previously unchar- diseases. Indeed, MPZL3 is less expressed in cSCC than in un- acterized new candidates (Tables S2 and S3) that could be differentiated subconfluent keratinocytes, and given its require- explored to better understand various aspects of somatic tissue ment for complete terminal differentiation, further work may differentiation. elucidate a mechanism of MPZL3 action in carcinogenesis. We also used a filter of opposite expression patterns in cSCC To validate the efficacy of proximity analysis and to better to generate the dynamic hub score that ranked proteins of inter- characterize the process of epidermal differentiation, we showed est. Given that this filter, alongside the parameters of small-world that MPZL3 is necessary for normal epidermal differentiation interaction connectedness, highlighted several genes with gene induction. The discovery that MPZL3 localizes to the

454 Developmental Cell 35, 444–457, November 23, 2015 ª2015 Elsevier Inc. mitochondria is consistent with the phenotypes that have been lenge and compared to networks based upon the supplemental material of observed in MPZL3 mutant and knockout mice, which Marbach et al. (2012). GO term annotations used for scoring algorithms was display lower body weight, even in the context of high fat diets, downloaded from the Saccharomyces Genome Database (Cherry et al., 2012). Analysis for epidermal differentiation was performed on data from the as well as lower blood glucose levels (Cao et al., 2007; Czyzyk GEO repository (GEO: GSE52651, GSE58161, and GSE35468) as well as a et al., 2013; Leiva et al., 2014). These previous observations, new microarray 7-day time course of keratinocytes seeded at confluence paired with its localization determined here, suggest MPZL3 with calcium. For this analysis, prior to the rank ordering step of analysis, the may have other important biological roles that would be of fold change of each gene was calculated between the dataset being analyzed interest to explore in future work. In addition to binding and the corresponding undifferentiated keratinocyte control. Clustering coef- FDXR, our mass spectrometry experiment identified a number ficients were calculated locally based upon the metrics previously described (Watts and Strogatz, 1998). Scripts used for proximity analysis and related of other mitochondrial proteins that are proximal to MPZL3, analyses are freely available for download at http://khavarilab.stanford.edu/ some of which have already been described as related to tools-1/#tools. one another metabolically or physically and are excellent candidates to pursue in better characterizing additional MPZL3 Dynamic Hub Score functions. The dynamic hub score was calculated from the local connectivity for MPZL3 and FDXR were found to bind one another and to regu- each node in the network, as measured by the number of ‘‘triangles’’ it was late ROS, with both MPZL3-FDXR binding and ROS induction involved in, of any size. This triangle number was weighted by the probability appearing necessary for normal differentiation. It is possible of the interactions assayed and finally weighted by the degree of inverse expression change observed in cSCC data (Arron et al., 2011). The top 100 that ROS plays a role in feedforward regulation that is necessary scores are provided in Table S3, excluding any genes not expressed in the to complete this process in a manner that is analogous to its role cSCC dataset. in p53 and FDXR-mediated apoptosis (Liu and Chen, 2002). Given that FDXR enzymatic function appears to be required for Cell Culture its functions in differentiation, this suggests that FDXR enzymatic Primary human keratinocytes were isolated from fresh, surgically discarded use of its substrate NADPH may prevent its quenching of ROS, neonatal foreskin. Keratinocytes were grown in Keratinocyte-SFM and allowing its accumulation and promotion of differentiation. In Medium 154 (Life Technologies) in equal proportions with included supple- this context, it is unclear how MPZL3 enables FDXR activity. It ments. Differentiation conditions consist of keratinocytes seeded to full confluence with 1.2 mM calcium for 3 days. 293T cells were grown in DMEM with is possible that MPZL3, as a transmembrane protein, acts as 10% fetal bovine serum. Organotypic culture was performed as previously an anchor to bind FDXR and increase its local concentration described (Truong et al., 2006). within the mitochondria to enhance its catalytic activity. Consis- tent with this possibility are the positive impacts on differentia- RNA Sequencing and Analysis tion of enforced FDXR expression. The increasing body of RNA sequencing libraries were prepared with TruSeqv2 library preparation kit literature suggesting that ROS is tightly regulated to control (Illumina) and sequenced with the Illumina HiSeq platform using 101-bp cell state transitions (Barbieri et al., 2003; Tormos et al., 2011; paired-end sequencing. Alignment was performed with TopHat2, and Cuffdiff2 Zhang et al., 2013) suggests that there may be similar mecha- was used to call differential gene expression. nisms regulating the expression of other proteins impacting ROS production, a possibility that warrants further study. Addi- Metabolomics Metabolomics analysis was performed in differentiated keratinocytes treated tional assays including examination of mitochondrial mor- with either control siRNA or independent siRNAs against MPZL3. These exper- phology and metabolomics suggest that there are roles for iments were performed on five biological replicates and were conducted by MPZL3 beyond the regulation of ROS described here and may Metabolon. Five mass spectrometry methods, including ultra performance be an interesting avenue for future investigation of MPZL3 and liquid chromatography-mass spectrometry (UPLC-MS), positive and negative mitochondria in epithelial tissue differentiation. ionization LC-MS, liquid chromatography polar mass spectrometry (LC-polar- MS), and gas chromatography-mass spectrometry (GC-MS), were used to identify 621 compounds in these keratinocyte cells. The complete dataset is EXPERIMENTAL PROCEDURES included in Table S5.

The Supplemental Experimental Procedures section, which includes addi- Statistical Methods tional experimental procedures, siRNA sequences, and qPCR primer se- All graphs with error bars are shown with either the standard deviation or stan- quences, is available online. dard error of the mean as indicated in the figure legends. The Student’s t test was used to calculate significance, unless otherwise indicated in the figure Proximity Analysis legend. Metabolomics statistics were performed using a one-way ANOVA Proximity analysis begins with isolating expressed transcripts in the sample of analysis and were performed by Metabolon. interest and rank ordering them from 0 to 1 with an even distribution across the range within each dataset. Next, the Pearson correlation is calculated between ACCESSION NUMBERS all gene pairs. Finally, using this correlation matrix as the input for topological constraints, with optimized metrics based upon DreamNetWeaver in silico The RNA sequencing and microarray data from this work have been deposited simulations, a random seed network is generated and nodes are added to in the Sequencing Read Archive with accession number PRJNA285272. this seed network if the absolute value of the correlation between the nodes being connected is greater than 0.5. This process continues until maximum limits have been met or all nodes have been tested, and it is iterated 10,000 SUPPLEMENTAL INFORMATION times to generate a final probability for any edge that was considered. A simpli- fied version of this iteration step was derived to minimize computational time Supplemental Information includes Supplemental Experimental Procedures, and both produced similar results; both versions are available online. Proximity six figures, and six tables and can be found with this article online at http:// analysis was performed on S. cerevisiae data from the DREAM5 network chal- dx.doi.org/10.1016/j.devcel.2015.10.023.

Developmental Cell 35, 444–457, November 23, 2015 ª2015 Elsevier Inc. 455 AUTHOR CONTRIBUTIONS Grabe, N., Pommerencke, T., Steinberg, T., Dickhaus, H., and Tomakidi, P. (2007). Reconstructing protein networks of epithelial differentiation from histo- Conceptualization, A.B., A.U., and P.A.K.; Methodology, A.B. and A.U.; Inves- logical sections. Bioinformatics 23, 3200–3208. tigation, A.B. and L.D.B.; Resources, A.B., V.L.-P., B.J.Z., and P.A.K.; Writing – Hamanaka, R.B., Glasauer, A., Hoover, P., Yang, S., Blatt, H., Mullen, A.R., Original Draft, A.B. and P.A.K.; Writing – Review & Editing, A.U., L.D.B, V.L.-P., Getsios, S., Gottardi, C.J., DeBerardinis, R.J., Lavker, R.M., and Chandel, and B.J.Z.; Funding Acquisition, P.A.K. and A.B.; Supervision, P.A.K. N.S. (2013). Mitochondrial reactive oxygen species promote epidermal differentiation and hair follicle development. Sci. Signal. 6, ra8. ACKNOWLEDGMENTS Herrga˚ rd, M.J., Swainston, N., Dobson, P., Dunn, W.B., Arga, K.Y., Arvas, M., Blu¨ thgen, N., Borger, S., Costenoble, R., Heinemann, M., et al. (2008). We thank H. Chang, T. Stearns, M. Teruel, and A. Oro, as well as B. Sun and X. A consensus yeast metabolic network reconstruction obtained from a commu- Bao, for pre-submission review. We thank J. Kovalski, B. Sun, and E. Turk for nity approach to systems biology. Nat. Biotechnol. 26, 1155–1160. helpful discussions and reagents, S. Tao and E. Obrien for technical assistance, and L. Morcom and P. Bernstein for administrative assistance. This Hopkin, A.S., Gordon, W., Klein, R.H., Espitia, F., Daily, K., Zeller, M., Baldi, P., work was supported by the U.S. Department of Veterans Affairs Office of and Andersen, B. (2012). GRHL3/GET1 and trithorax group members collabo- Research and Development, by NIAMS NIH R01 AR45192 (P.A.K), and by rate to activate the epidermal progenitor differentiation program. PLoS Genet. NCI F310CA180408 (A.B.). 8, e1002829. Hughes, T.R., Marton, M.J., Jones, A.R., Roberts, C.J., Stoughton, R., Armour, Received: May 29, 2015 C.D., Bennett, H.A., Coffey, E., Dai, H., He, Y.D., et al. (2000). Functional dis- Revised: October 19, 2015 covery via a compendium of expression profiles. Cell 102, 109–126. Accepted: October 26, 2015 Hwang, P.M., Bunz, F., Yu, J., Rago, C., Chan, T.A., Murphy, M.P., Kelso, G.F., Published: November 23, 2015 Smith, R.A., Kinzler, K.W., and Vogelstein, B. (2001). Ferredoxin reductase af- fects p53-dependent, 5-fluorouracil-induced apoptosis in colorectal cancer REFERENCES cells. Nat. Med. 7, 1111–1117. Kasahara, A., and Scorrano, L. (2014). Mitochondria: from cell death execu- Arron, S.T., Ruby, J.G., Dybbro, E., Ganem, D., and Derisi, J.L. (2011). tioners to regulators of cell differentiation. Trends Cell Biol. 24, 761–770. Transcriptome sequencing demonstrates that human papillomavirus is not Kretz, M., Webster, D.E., Flockhart, R.J., Lee, C.S., Zehnder, A., Lopez- active in cutaneous squamous cell carcinoma. J. Invest. Dermatol. 131, Pajares, V., Qu, K., Zheng, G.X., Chow, J., Kim, G.E., et al. (2012). 1745–1753. Suppression of progenitor differentiation requires the long noncoding RNA Barabasi, A.L., and Albert, R. (1999). Emergence of scaling in random net- ANCR. Genes Dev. 26, 338–343. works. Science 286, 509–512. Kretz, M., Siprashvili, Z., Chu, C., Webster, D.E., Zehnder, A., Qu, K., Lee, C.S., Barbieri, S.S., Eligini, S., Brambilla, M., Tremoli, E., and Colli, S. (2003). Flockhart, R.J., Groff, A.F., Chow, J., et al. (2013). Control of somatic tissue dif- Reactive oxygen species mediate cyclooxygenase-2 induction during mono- ferentiation by the long non-coding RNA TINCR. Nature 493, 231–235. cyte to macrophage differentiation: critical role of NADPH oxidase. Kwa, R.E., Campana, K., and Moy, R.L. (1992). Biology of cutaneous squa- Cardiovasc. Res. 60, 187–197. mous cell carcinoma. J. Am. Acad. Dermatol. 26, 1–26. Boxer, L.D., Barajas, B., Tao, S., Zhang, J., and Khavari, P.A. (2014). ZNF750 interacts with KLF4 and RCOR1, KDM1A, and CTBP1/2 chromatin regulators Lambeth, J.D., and Kamin, H. (1976). Adrenodoxin reductase. Properties of the to repress epidermal progenitor genes and induce differentiation genes. complexes of reduced enzyme with NADP+ and NADPH. J. Biol. Chem. 251, Genes Dev. 28, 2013–2026. 4299–4306. Cao, T., Racz, P., Szauter, K.M., Groma, G., Nakamatsu, G.Y., Fogelgren, B., Leiva, A.G., Chen, A.L., Devarajan, P., Chen, Z., Damanpour, S., Hall, J.A., Pankotai, E., He, Q.P., and Csiszar, K. (2007). Mutation in Mpzl3, a novel [cor- Bianco, A.C., Li, J., Badiavas, E.V., Zaias, J., et al. (2014). Loss of Mpzl3 func- rected] gene encoding a predicted [corrected] adhesion protein, in the rough tion causes various skin abnormalities and greatly reduced adipose depots. coat (rc) mice with severe skin and hair abnormalities. J. Invest. Dermatol. J. Invest. Dermatol. 134, 1817–1827. 127, 1375–1386. Liu, G., and Chen, X. (2002). The ferredoxin reductase gene is regulated by the Carro, M.S., Lim, W.K., Alvarez, M.J., Bollo, R.J., Zhao, X., Snyder, E.Y., p53 family and sensitizes cells to oxidative stress-induced apoptosis. Sulman, E.P., Anne, S.L., Doetsch, F., Colman, H., et al. (2010). The transcrip- Oncogene 21, 7195–7204. tional network for mesenchymal transformation of brain tumours. Nature 463, Lopez-Pajares, V., Qu, K., Zhang, J., Webster, D.E., Barajas, B.C., Siprashvili, 318–325. Z., Zarnegar, B.J., Boxer, L.D., Rios, E.J., Tao, S., et al. (2015). A LncRNA- Chagoyen, M., and Pazos, F. (2010). Quantifying the biological significance of MAF:MAFB transcription factor network regulates epidermal differentiation. gene ontology biological processes–implications for the analysis of systems- Dev. Cell 32, 693–706. wide data. Bioinformatics 26, 378–384. Marbach, D., Costello, J.C., Ku¨ ffner, R., Vega, N.M., Prill, R.J., Camacho, Cherry, J.M., Hong, E.L., Amundsen, C., Balakrishnan, R., Binkley, G., Chan, D.M., Allison, K.R., Kellis, M., Collins, J.J., and Stolovitzky, G.; DREAM5 E.T., Christie, K.R., Costanzo, M.C., Dwight, S.S., Engel, S.R., et al. (2012). Consortium (2012). Wisdom of crowds for robust gene network inference. Saccharomyces Genome Database: the genomics resource of budding yeast. Nat. Methods 9, 796–804. Nucleic Acids Res. 40, D700–D705. Mellacheruvu, D., Wright, Z., Couzens, A.L., Lambert, J.P., St-Denis, N.A., Li, Choi, H., Larsen, B., Lin, Z.Y., Breitkreutz, A., Mellacheruvu, D., Fermin, D., T., Miteva, Y.V., Hauri, S., Sardiu, M.E., Low, T.Y., et al. (2013). The CRAPome: Qin, Z.S., Tyers, M., Gingras, A.C., and Nesvizhskii, A.I. (2011). SAINT: proba- a contaminant repository for affinity purification-mass spectrometry data. Nat. bilistic scoring of affinity purification-mass spectrometry data. Nat. Methods 8, Methods 10, 730–736. 70–73. Mills, A.A., Zheng, B., Wang, X.J., Vogel, H., Roop, D.R., and Bradley, A. Czyzyk, T.A., Andrews, J.L., Coskun, T., Wade, M.R., Hawkins, E.D., (1999). p63 is a p53 homologue required for limb and epidermal morphogen- Lockwood, J.F., Varga, G., Sahr, A.E., Chen, Y., Brozinick, J.T., et al. (2013). esis. Nature 398, 708–713. Genetic ablation of myelin protein zero-like 3 in mice increases energy expen- Mootha, V.K., Lindgren, C.M., Eriksson, K.F., Subramanian, A., Sihag, S., diture, improves glycemic control, and reduces hepatic lipid synthesis. Am. J. Lehar, J., Puigserver, P., Carlsson, E., Ridderstra˚ le, M., Laurila, E., et al. Physiol. Endocrinol. Metab. 305, E282–E292. (2003). PGC-1alpha-responsive genes involved in oxidative phosphorylation Feizi, S., Marbach, D., Me´ dard, M., and Kellis, M. (2013). Network deconvolu- are coordinately downregulated in human diabetes. Nat. Genet. 34, 267–273. tion as a general method to distinguish direct dependencies in networks. Nat. Patel, S., Xi, Z.F., Seo, E.Y., McGaughey, D., and Segre, J.A. (2006). Klf4 and Biotechnol. 31, 726–733. corticosteroids activate an overlapping set of transcriptional targets to

456 Developmental Cell 35, 444–457, November 23, 2015 ª2015 Elsevier Inc. accelerate in utero epidermal barrier acquisition. Proc. Natl. Acad. Sci. USA Teng, A., Nair, M., Wells, J., Segre, J.A., and Dai, X. (2007). Strain-dependent 103, 18668–18673. perinatal lethality of Ovol1-deficient mice and identification of Ovol2 as a Pichiorri, F., Okumura, H., Nakamura, T., Garrison, P.N., Gasparini, P., Suh, downstream target of Ovol1 in skin epidermis. Biochim. Biophys. Acta 1772, S.S., Druck, T., McCorkell, K.A., Barnes, L.D., Croce, C.M., and Huebner, K. 89–95. (2009). Correlation of fragile histidine triad (Fhit) protein structural features Tong, A.H., Lesage, G., Bader, G.D., Ding, H., Xu, H., Xin, X., Young, J., Berriz, with effector interactions and biological functions. J. Biol. Chem. 284, 1040– G.F., Brost, R.L., Chang, M., et al. (2004). Global mapping of the yeast genetic 1049. interaction network. Science 303, 808–813. Racz, P., Mink, M., Ordas, A., Cao, T., Szalma, S., Szauter, K.M., and Csiszar, Tormos, K.V., Anso, E., Hamanaka, R.B., Eisenbart, J., Joseph, J., K. (2009). The human orthologue of murine Mpzl3 with predicted adhesive and Kalyanaraman, B., and Chandel, N.S. (2011). Mitochondrial complex III ROS immune functions is a potential candidate gene for immune-related hereditary regulate adipocyte differentiation. Cell Metab. 14, 537–544. hair loss. Exp. Dermatol. 18, 261–263. Truong, A.B., Kretz, M., Ridky, T.W., Kimmel, R., and Khavari, P.A. (2006). p63 Ratushny, V., Gober, M.D., Hick, R., Ridky, T.W., and Seykora, J.T. (2012). regulates proliferation and differentiation of developmentally mature keratino- From keratinocyte to cancer: the pathogenesis and modeling of cutaneous cytes. Genes Dev. 20, 3185–3197. squamous cell carcinoma. J. Clin. Invest. 122, 464–472. UniProt Consortium (2015). UniProt: a hub for protein information. Nucleic Ravasz, E., Somera, A.L., Mongru, D.A., Oltvai, Z.N., and Baraba´ si, A.L. (2002). Acids Res. 43, D204–D212. Hierarchical organization of modularity in metabolic networks. Science 297, van Noort, V., Snel, B., and Huynen, M.A. (2004). The yeast coexpression 1551–1555. network has a small-world, scale-free architecture and can be explained by Roux, K.J., Kim, D.I., Raida, M., and Burke, B. (2012). A promiscuous biotin a simple model. EMBO Rep. 5, 280–284. ligase fusion protein identifies proximal and interacting proteins in mammalian Watts, D.J., and Strogatz, S.H. (1998). Collective dynamics of ‘small-world’ cells. J. Cell Biol. 196, 801–810. networks. Nature 393, 440–442. Rowe, D.E., Carroll, R.J., and Day, C.L., Jr. (1992). Prognostic factors for local Xu, X., Duan, S., Yi, F., Ocampo, A., Liu, G.H., and Izpisua Belmonte, J.C. recurrence, metastasis, and survival rates in squamous cell carcinoma of the (2013). Mitochondrial regulation in pluripotent stem cells. Cell Metab. 18, skin, ear, and lip. Implications for treatment modality selection. J. Am. Acad. 325–332. Dermatol. 26, 976–990. Xue, S., and Barna, M. (2012). Specialized ribosomes: a new frontier in gene Schaffter, T., Marbach, D., and Floreano, D. (2011). GeneNetWeaver: in silico regulation and organismal biology. Nat. Rev. Mol. Cell Biol. 13, 355–369. benchmark generation and performance profiling of network inference Yang, A., Schweitzer, R., Sun, D., Kaghad, M., Walker, N., Bronson, R.T., methods. Bioinformatics 27, 2263–2270. Tabin, C., Sharpe, A., Caput, D., Crum, C., and McKeon, F. (1999). p63 is Segre, J.A., Bauer, C., and Fuchs, E. (1999). Klf4 is a transcription factor essential for regenerative proliferation in limb, craniofacial and epithelial devel- required for establishing the barrier function of the skin. Nat. Genet. 22, opment. Nature 398, 714–718. 356–360. Yang, D., Sun, Y., Hu, L., Zheng, H., Ji, P., Pecot, C.V., Zhao, Y., Reynolds, S., Sen, G.L., Boxer, L.D., Webster, D.E., Bussat, R.T., Qu, K., Zarnegar, B.J., Cheng, H., Rupaimoole, R., et al. (2013). Integrated analyses identify a master Johnston, D., Siprashvili, Z., and Khavari, P.A. (2012). ZNF750 is a p63 target microRNA regulatory network for the mesenchymal subtype in serous ovarian gene that induces KLF4 to drive terminal epidermal differentiation. Dev. Cell cancer. Cancer Cell 23, 186–199. 22, 669–677. Yook, S.H., Oltvai, Z.N., and Baraba´ si, A.L. (2004). Functional and topological Settembre, C., Fraldi, A., Medina, D.L., and Ballabio, A. (2013). Signals from characterization of protein interaction networks. Proteomics 4, 928–942. the lysosome: a control centre for cellular clearance and energy metabolism. Yosef, N., Shalek, A.K., Gaublomme, J.T., Jin, H., Lee, Y., Awasthi, A., Wu, C., Nat. Rev. Mol. Cell Biol. 14, 283–296. Karwacz, K., Xiao, S., Jorgolli, M., et al. (2013). Dynamic regulatory network So¨ derberg, O., Leuchowius, K.J., Gullberg, M., Jarvius, M., Weibrecht, I., controlling TH17 cell differentiation. Nature 496, 461–468. Larsson, L.G., and Landegren, U. (2008). Characterizing proteins and their Yu, Z., Lin, K.K., Bhandari, A., Spencer, J.A., Xu, X., Wang, N., Lu, Z., Gill, G.N., interactions in cells and tissues using the in situ proximity ligation assay. Roop, D.R., Wertz, P., and Andersen, B. (2006). The Grainyhead-like epithelial Methods 45, 227–232. transactivator Get-1/Grhl3 regulates epidermal terminal differentiation and in- Suzuki, H., Forrest, A.R., van Nimwegen, E., Daub, C.O., Balwierz, P.J., Irvine, teracts functionally with LMO4. Dev. Biol. 299, 122–136. K.M., Lassmann, T., Ravasi, T., Hasegawa, Y., de Hoon, M.J., et al.; FANTOM Zhang, Y., Choksi, S., Chen, K., Pobezinskaya, Y., Linnoila, I., and Liu, Z.G. Consortium; Riken Omics Science Center (2009). The transcriptional network (2013). ROS play a critical role in the differentiation of alternatively activated that controls growth arrest and differentiation in a human myeloid leukemia cell macrophages and the occurrence of tumor-associated macrophages. Cell line. Nat. Genet. 41, 553–562. Res. 23, 898–914.

Developmental Cell 35, 444–457, November 23, 2015 ª2015 Elsevier Inc. 457